Experiment 3: Oxygen Generator ; (pre-setup)


This is another experiment principally aimed at improving my life and overall physical performance somehow.  Obviously experiments on one’s self are subject to huge amounts of bias, which is why I am not assessing (yet) the “benefit” (or placebo-nefit) of my experimental results.  The motivation though for this particular experiment comes from the observation that when I am able to breath and achieve full airspeed through my face, that I sleep better, have huge amounts of energy, burn more, am more intelligent (like 10x), and am generally more optimistic about the future.  Since I am still waiting for the second procedure on my sinuses to take place and “fix” my breathing in the first order, I figured that another way to approach the solution is to drive up the oxygen mole fraction of oxygen in my bedroom, perhaps 10% or so.  (Much lower than the concentration needed to reach Oxygen Toxicity)


There are plenty of ways to “generate” (release) oxygen, and the way that I will be focusing on is at its core, nothing more than “simple electrolysis” of water (H2O).  Of course,  “simple” electrolysis isn’t quite so simple and there are plenty of ways to screw it up and end up with either no H2 or no O2, or neither, or worse, something different (say Chlorine, if you use table salt as an electrolyte).  Since I want to actually be breathing what comes out here, no Cl for me, if you please.  Anyway, that isn’t quite relevant at this stage, what we really want to know is if this experiment is even tractably possible.  i.e. under “ideal” (whatever that means) conditions, can enough O2 be produced to satisfy my airy needs?

My room is 12′x12′, with an average 15′ ceiling. That means that I can count on 452.5 cubic ft of O2 assuming 20.95% O2 by volume. A 10% increase should yield ~45cubic ft, or 1281 liters of O2. At 70F (instead of 0C) and 1 atmosphere of pressure, an ideal gas will occupy ~7% more space (294K/273K), so 1 mole of O2 will occupy 24.1 liters instead of 22.4L. 1281L/24.1L => 53.15 Moles of O2 (or ~106moles of just O) need to be released. So can I produce that much at a reasonable rate from electrolysis of water??

Well, first of all we know that 1Liter of water is 55.56 moles H2O, which means that I only have to dissociate 1.91L of water or “just” 1/2 gallon of water.

So we know that it can be “reasonably” done in pure volume terms, but what about electrically?

The electrolysis of water requires 237 kJ/mol of water, or 25,122 kJ to convert 106moles of H2O to 53 moles of O2. If I happen to have a power supply that outputs 120W, then 120W*x sec = 25,122 kJ => 209s, or 3.5 minutes. Now this sounds wrong, but the reason is because we are a) assuming 100% efficiency, and b) assuming that all 120W can actually be put into the water easily. Since I actually have a 20V P/S @ 6A, and electrolysis of water uses 1.2V, basically I am only putting 7.2W into the water. So at max efficiency I am looking at 58.15 minutes instead. This still seems too good to be true, but you get the picture, it is totally “possible”.


Now we come to the pre-setup experiments, basically where I test out the system and see if something happens, and identify what is (if anything) going wrong.

Mini Test 1:

  1. Pure Graphite Rods
  2. Distilled Water (“pure” H2O has a high dielectric strength and is actually an insulator)
  3. 2tbsp of Sodium Bicarbonate as an electrolyte

Mini Result 1:

Carbon anode gets totally corroded, dispersing carbon into the water-suspension.  Oxygen reacts with the anode as well, producing CO2, which dissolves again into the water.  Overall reduced (barely any) oxygen production.  Lots of H2 produced.

Mini Test 2:

Since I’m a cheapskate, is there enough platinum in the platinum-cobalt-chromium alloy plated onto hard-disk platters to keep the anode sufficiently inert so that oxygen can be produced without reacting with either the electrolyte or the anode?  Cuz using those would be cool (and I have like hundreds of unused drives). Those are the photos of me taking apart the hard drive.  Eventually I gave up trying to remove the platter and just immersed the entire disk into the solution to see in an hour or so if the platter was corroded or not, when acting as the anode.

Mini Result 2:

It ran for a while looking all shiny, but then as you can see, the platter is corroded.  So no, looks like I’ll actually have to get a platinum anode. 


Several lessons have been learned so far, for the development of the full O2 generator.

  1. platinum anode
  2. H2SO4 as electrolyte
  3. Multiple Anode+Cathode in series — specifically 16 pairs are needed to use the entire 20V drop from the power supply.
  4. Glass tank to handle the electrolyte acidity

Time for bed…

(Pic 1)

(Pic 2)

(Pic 3)

(Pic 4)

(Pic 5)

(Pic 6)

(Pic 7)